Description

The analytical methods most commonly used in clinical laboratories and for which harmonization/standardization of procedures is required are immunoassays, which enable the quantification of an analyte.

Currently, the reference method for immunoassays is liquid chromatography tandem mass spectrometry (LC-MS/MS). LC-MS/MS is increasingly used in clinical laboratories. Its high specificity, sensitivity, and multi-analyte capability make it an ideal alternative to immunoassays or conventional high-performance liquid chromatography (HPLC). LC-MS/MS also provides greater flexibility than immunoassays, as LC-MS/MS assays are typically developed in-house. In addition, a single LC-MS/MS run can generate a large amount of quantitative or qualitative data.

All of these methods are essential in clinical diagnostics because they allow the detection and quantification of a wide range of biomarkers, including hormones, proteins, and antibodies across numerous areas of medicine. Despite their widespread use and importance, variability between different methods, within the same laboratory or across laboratories, represents a significant challenge, with considerable impact on results, clinical decisions, and ultimately on patient care. This variability may depend on various factors, including the quality and source of reagents, the quality of calibrators critical for analyte quantification, detection methods, and testing protocols.

Every clinical laboratory is therefore required to assess the agreement of results whenever new instruments or analytical procedures are introduced, compared with a method already in use. Method comparison is indeed one of the principals of the International Standard (ISO 15189:2012) for the accreditation of medical laboratories and of the IVD Directive 98/79 EC, which defines the requirements for in vitro diagnostic devices (IVDs) at the European level.

The comparison of two methods requires defining a maximum acceptable difference between the concentrations of an analyte obtained using the analytical techniques of interest; therefore, two methods may be considered comparable if the differences between the results are below a predefined acceptable limit: the maximum acceptable error. The definition of the maximum acceptable error is based on analytical quality specifications established according to five hierarchical models defined at the 1999 Stockholm Conference:

• a model based on the evaluation of the effect of analytical performance on clinical outcomes under specific clinical conditions;
• a model based on the evaluation of the effect of analytical performance on clinical decisions;
• a model based on the biological variability of the analyte;
• a model based on published expert recommendations;
• a model based on the current state of the art (of instruments and/or methods), derived for example from EQA programs.

Therefore, the study will be structured into several phases, beginning with the collection of information on the analytical characteristics and performance of the instruments being compared, verifying method repeatability. For each analyte, the maximum acceptable error will be defined a priori, based on analytical quality specifications, such as analytical imprecision and inaccuracy.

Statistical tests will be performed to assess the error between the analyte measurements obtained using the different methods and, if comparability is not achieved, additional evaluations will be carried out to rule out methodological errors.